Skip to Main content Skip to Navigation
Conference papers

Log(lambda) Modifications for Optimal Parallelism

Fabien Teytaud 1, 2 Olivier Teytaud 1, 2
1 TAO - Machine Learning and Optimisation
CNRS - Centre National de la Recherche Scientifique : UMR8623, Inria Saclay - Ile de France, UP11 - Université Paris-Sud - Paris 11, LRI - Laboratoire de Recherche en Informatique
Abstract : It is usually considered that evolutionary algorithms are highly parallel. In fact, the theoretical speed-ups for parallel optimization are far better than empirical results; this suggests that evolutionary algorithms, for large numbers of processors, are not so efficient. In this paper, we show that in many cases automatic parallelization provably provides better results than the standard parallelization consisting of simply increasing the population size lambda. A corollary of these results is that logarithmic bounds on the speed-up (as a function of the number of computing units) are tight within constant factors. Importantly, we propose a simple modification, termed log(lambda)-correction, which strongly improves several important algorithms when lambda is large.
Complete list of metadatas

Cited literature [14 references]  Display  Hide  Download

https://hal.inria.fr/inria-00495087
Contributor : Fabien Teytaud <>
Submitted on : Friday, June 25, 2010 - 9:02:21 AM
Last modification on : Wednesday, September 16, 2020 - 5:07:18 PM
Long-term archiving on: : Monday, September 27, 2010 - 11:56:04 AM

File

autoparacnf.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00495087, version 1

Collections

Citation

Fabien Teytaud, Olivier Teytaud. Log(lambda) Modifications for Optimal Parallelism. Parallel Problem Solving From Nature, Sep 2010, Krakow, Poland. ⟨inria-00495087⟩

Share

Metrics

Record views

350

Files downloads

393